Double-deck elevator systems using genetic network programming based on variance information

Double-deck elevator systems (DDES) have been invented to improve the transportation capacity of elevator group systems for decades. There are several specific features in DDES due to its specific structure, i.e., two decks are vertically connected in one shaft. Even though the DDES could work well in a pure up-peak traffic pattern by cutting up to half of the stops in an elevator round trip, it becomes intractable because of the features when running in some other traffic patterns. Some solutions employing evolutionary computation methods such as genetic algorithm were also proposed in recent years. In this paper, we propose an approach of DDES using genetic network programming based on our past studies in this field.